The financial sector is under increasing pressure to outpace fraudsters who constantly adapt to new technologies. In the latest episode of Siili's Composable Banking Podcast, host Noora Teronen spoke with Sara Helin, a fraud prevention expert at Nordea, about how artificial intelligence (AI) helps combat fraud detection in the Nordic banking sector.
Sara shares insights into banking fraud and the role of AI in detecting fraudulent transactions, and how financial institutions must balance security with customer convenience while staying ahead of emerging threats.
Key talking points from the discussion
✅ AI boosts fraud detection by automating processes and handling false positives.
✅ Balancing security and customer convenience is crucial—too much friction can damage trust.
✅ AI helps fraud teams prioritize resources by focusing on real threats instead of false alarms.
✅ Fraud tactics constantly evolve, and banks must stay ahead through AI adaptation.
✅ The future of AI in fraud prevention depends on better data, improved infrastructure, and collaboration.
AI in fraud detection: From manual efforts to automation
Fraud detection has traditionally relied on manual review processes and rule-based models, which, while effective to an extent, have struggled to keep pace with the scale and sophistication of modern fraud tactics. With millions of transactions occurring daily, human-led detection alone is insufficient to effectively identify threats in real-time.
AI is changing that.
Sara highlights how AI enhances fraud detection capabilities by automating processes, allowing banks to analyze massive amounts of transaction data quickly and efficiently. By learning from historical fraud cases, AI adapts dynamically, identifying fraudulent behavior with increasing accuracy over time.
Another key challenge that AI helps address is the high rate of false positives in fraud detection systems. Traditional approaches often block legitimate transactions, frustrating customers and increasing operational burdens. AI, however, refines fraud detection by distinguishing between genuine and suspicious activities with greater precision.
Yet, as AI becomes more integrated into fraud prevention strategies, banks must carefully consider the balance between security and convenience.
Balancing security with seamless customer experiences
One of the biggest pain points in fraud prevention is striking the right balance between robust security and a smooth customer experience. Customers expect frictionless banking, where payments are instant and hassle-free—but at the same time, security controls must be in place to protect against fraud.
Sara emphasizes this constant balancing act: "We also need to constantly think about how much friction we can have when people are making payments."
On one hand, adding too many security checks—such as additional authentication layers—can make banking frustrating for users. On the other hand, reducing security checks too much can increase fraud risks, exposing customers to potential financial losses.
This is where AI plays a critical role. AI-driven fraud detection can operate in the background, analyzing transactions without disrupting the user experience. AI learns each customer’s behavior, identifying anomalies without requiring constant manual intervention.
Sara explains that fraud prevention teams must constantly assess security controls to ensure they do not introduce unnecessary obstacles for customers: "It needs to be like instantly. Exactly. So, so that's really also important piece of the puzzle and how to kind of tackle that, that everything needs to happen really quickly. And then you would need to also do those checks that is everything okay with this. That is this fraud or not. So it's definitely, definitely a challenge in this area."
Prioritizing human expertise where it matters
AI does not just help detect fraud—it also improves operational efficiency within banks, enabling fraud prevention teams to focus on high-risk cases rather than being overwhelmed with thousands of potential fraud alerts. Sara highlights how AI is helping fraud teams work smarter: "With AI, we can concentrate our resources where they are most needed. Instead of spending time on false alarms, we can focus on real threats."
By automating routine fraud checks, AI frees up human investigators to focus on more complex and high-value fraud cases, where manual review is essential. Furthermore, AI’s ability to detect patterns in fraudulent transactions helps banks refine their fraud prevention strategies over time.
Sara points out that fraud prevention is not just about stopping fraud—it’s also about continuous improvement and adaptation: "Fraud is ever-evolving. The methods fraudsters use today are different from what they were using just a year ago. It’s a constant game of cat and mouse."
Fraud trends in the Nordics: A growing challenge
Nordic banks are not immune to fraud, and Sara highlights some of the most common fraud tactics used in the region.
One of the most prevalent fraud types in Finland is phishing: "We have also quite a lot of phishing activities in Finland, which is something that we don't see in other Nordic countries. So for some reason in Finland, the kind of phishing in search engines is really heavy."
These phishing scams often target customers via SMS, emails, and fraudulent websites, tricking them into revealing banking credentials. More sophisticated fraud tactics are also emerging, including AI-generated scams that make phishing attacks even more convincing.
Sara warns that fraudsters are now leveraging AI for their own benefit: "It just boosts like the existing fraud types that we have and fraud modus operandis. For example, in love scams, you can create like video material of a person that doesn't even exist or pictures. In investment scams, you can also utilize AI in very many ways."
She also highlights that criminals are using AI-powered language tools to create more realistic phishing messages in local languages, making it harder for people to identify scams: "Basically, anyone can create a Finnish-speaking text quite well already. It might not be perfect, but it's perfect enough to fool people, really."
Looking ahead: The future of AI in fraud prevention
As AI continues to evolve, its role in fraud prevention will only become more significant. However, for AI to reach its full potential, banks must address challenges related to data quality and integration.
Sara highlights the importance of data infrastructure: "To utilize AI, you have to have proper data and proper processes to utilize it in full. So I think maybe banks have a lot to do with that as well."
She also sees AI as a critical enabler for improving fraud detection while maintaining customer trust: "The best security is the security that you don’t see and experience. So many things are happening in the background all the time, and that’s the best thing if you never notice it."
Sara Helin (LinkedIn) is a fraud prevention expert at Nordea, specializing in fraud detection, AI-driven prevention strategies, and financial security. She investigates fraud trends, supports fraud case investigations, and helps optimize fraud detection systems.
Noora Teronen (LinkedIn) is a Banking Vodcast Host and Senior Sales Manager at Siili Solutions, focusing on AI in banking, fraud prevention, and digital transformation in financial services.